lable at ScienceDirect
Learning and Instruction 35 (2015) 94e103
Contents lists avai
Learning and Instruction
journal homepage: www.elsevier .com/locate/ learninstruc
Effects of a classroom discourse intervention on teachers' practice andstudents' motivation to learn mathematics and science
Katharina Kiemer*, Alexander Gr€oschner 1, Ann-Kathrin Pehmer 2, Tina Seidel 3
Technische Universit€at München, Arcisstraße 21, 80333 Munich, Germany
a r t i c l e i n f o
Article history:Received 15 July 2014Received in revised form29 September 2014Accepted 22 October 2014Available online
Keywords:Teacher professional developmentClassroom discourseSelf-determination theoryInterestIntervention
* Corresponding author. Tel.: þ49 89 289 25124; faE-mail addresses: [email protected] (K. Kie
tum.de (A. Gr€oschner), ann-kathrin.pehmer@[email protected] (T. Seidel).
1 Tel.: þ49 89 289 25121.2 Tel.: þ49 89 289 25123.3 Tel.: þ49 89 289 25114.
http://dx.doi.org/10.1016/j.learninstruc.2014.10.0030959-4752/© 2014 Elsevier Ltd. All rights reserved.
a b s t r a c t
Student interest and motivation in STEM subjects has dropped significantly throughout secondary ed-ucation, for which teacherestudent interactions are named as a central reason. This study investigatedwhether a video-based teacher professional development (TPD) intervention on productive classroomdiscourse improved students' learning motivation and interest development over the course of a schoolyear. The teachers' intervention group (IG; n ¼ 6) was compared with a control group (CG; n ¼ 4) whoparticipated in a traditional TPD programme on classroom discourse. The teachers showed a significantincrease in constructive feedback and decrease in simple feedback as a function of the treatment. Pre-and post-tests revealed that students in the IG (n ¼ 136) significantly increased their perceived auton-omy, competence and intrinsic learning motivation as compared with those in the CG (n ¼ 90). They alsoshowed significantly greater interest changes in the subjects compared with their peers in the CG.
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
Motivational concepts such as interest in the subject areimportant outcomes of educational processes (Krapp & Prenzel,2011) and are key elements regarding the young generations' pre-paredness for life-long learning as a core-skill in knowledge-basedsocieties. Motivation and interest development, especially in sci-ence, technology, engineering and mathematic (STEM) subjects,also determine adolescents' willingness to choose STEM-relatedcareer paths (Organisation for Economic Co-operation and Devel-opment, 2007).
Therefore, developing students' interest has to be a maineducational objective for schools as well as individual teachers.Interested learners develop more differentiated domain-specificknowledge (Renninger, Hidi, & Krapp, 1992), are more focusedand have better attention (Ainley, Hidi, & Berndorff, 2002), pursuemastery rather than performance goals (Harackiewicz, Durik,Barron, & Linnenbrink, 2008) and receive better grades than
x: þ49 89 289 25199.mer), [email protected] (A.-K. Pehmer), tina.
uninterested learners (Schiefele, Krapp, & Winteler, 1992). How-ever, interest in the subjects decreases significantly throughoutsecondary education (Eccles et al., 1993; Maulana, Opdenakker, &den Brok, submitted for publication). One reason frequently citedfor this decrease is the mismatch between students' needs andclassroom practices, especially during secondary education (Eccleset al., 1993).
For positive outcomes of motivated learning, it is the quality ofmotivation that is decisive (Ryan&Deci, 2000). Interest-based, self-determined forms of learning motivation provide the mostfavourable learning outcomes (Krapp, 2002; Krapp & Ryan, 2002).Yet, classroom interactions are often dominated by close-formattedclassroom discourse (Jurik, Gr€oschner, & Seidel, 2013; Walshaw &Anthony, 2008; Wells & Arauz, 2006), with negative effects bothon students' situational experiences of self-determined learningmotivation and their long-term development of more stable ori-entations (e.g. interest in STEM subjects).
Despite debates about enriching teaching through diversedidactical methods, classroom discourse and verbal teachere-student interactions remain dominant in mathematics and scienceeducation (Roth et al., 2006). Typical for practices of classroomdiscourse is a questioning-developing teaching style in which theteacher dominantly steers the interactions. However, if teacherscan actively engage their students in classroom discourse, they arelikely to engage them in more meaningful and sustained learningexperiences (Walshaw & Anthony, 2008). Therefore, changing
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103 95
students' classroom experiences into meaningful ones that developtheir interests by changing the prominent routines of classroominteractions and discourse is important. Productive classroomdiscourse is a key aspect for students' development of interest inthe subject.
To improve productive classroom discourse, video examples ofclassroom interactions are a promising tool for supporting teachers'analysis and reflection on classroom practices and interactionpatterns (Tripp & Rich, 2012). Using video in TPD facilitatesteachers' (emotional) involvement and makes them more investedin the learning process by activating prior knowledge, buildingpractical knowledge and bridging the theoryepractice gap(Santagata & Guarino, 2010).
1.1. Theoretical background
This study focuses on a video-based teacher professionaldevelopment (TPD) intervention to foster teachers' skills leading toproductive classroom discourse, which is compared with tradi-tional TPD in the German context. We investigated the extent towhich a newly developed intervention (Gr€oschner, Seidel, Kiemer,& Pehmer, 2014) positively affected teachers' practices as well asstudents' experiences of self-determined learning motivation andtheir development of interest in the subject. Thus, this study shedslight on the benefits of video-based TPD in comparison to morecommon practices of professional development in the Germancontext (Richter, Kunter, Klusmann, Lüdtke, & Baumert, 2011), aswell as on the importance of meaningful classroom discourse(specifically, teacher questioning and feedback) for students' situ-ational and long-term motivational orientations.
1.1.1. Classroom discourse and the role of questioning and feedbackVerbal teacherestudent and peer interactions are major means
to construct meaning (Mercer, 2010; Oliveira, 2010; Webb, 2009).Language use and interaction quality have important implicationsfor students' learning processes and outcomes (Lipowsky, Rakoczy,Pauli, Reusser, & Klieme, 2007), active engagement, learningmotivation and interest (Sierens, Vansteenkiste, Goossens,Soenens, & Dochy, 2009).
Walshaw and Anthony (2008) differentiate two major teachingstrategies in productive classroom discourse: clarifying discourseparticipation rights and responsibilities between the teacher andstudents, where the objective is to engage students in classroomconversation, and scaffolding students' ideas, e.g. by giving individualfeedback, in a productive way to move thinking forward. Severalstudies on mathematical argumentation and science inquiry pointout the relevance of these activities in creating productive class-room discourse (Furtak, Seidel, Iverson, & Briggs, 2012; Kovolainen& Kumpulainen, 2005). The two activities can particularly be con-ceptualised through productive forms of teacher questioning andmeaningful feedback (Jurik, Gr€oschner, & Seidel, 2014).
Teacher questioning: Meaningful patterns of teacher questioningleave room for students to explore and express their own under-standing. Generally, questioning strategies such as using openquestions have been shown to support student motivation andengagement through verbal discourse (Jurik et al., 2014).
Teacher feedback: Teacher feedback is deemed supportive oflearning and motivation, not just when it informs students aboutthe correctness (‘yes’, ‘right’) of a response, but especially when itincludes information about what aspects of the response are cor-rect (‘You did a good job on that graph’) or incorrect and how anyflaws can be mended (‘Try concentrating more closely on the transferof electrons’) or generally supports the learning process (‘No prob-lem.Wewill get everyone to understand that concept today’) (Hattie&Timperley, 2007).
TPD programmes focussing on productive classroom discoursein STEM subjects are rare. By testing the effectiveness of an inno-vative TPD programme including its effects on students' motivationand interest (Desimone, 2009; Wilson, 2013), the present studybreaks new ground in this line of research.
1.1.2. Classroom discourse and learning motivationTo investigate whether an innovative TPD programme focussing
on productive classroom discourse can affect students' motivationto learn, we draw on self-determination theory (Deci & Ryan, 1985,2004) and personeobject theory of interest (Krapp, 2002; Krapp &Prenzel, 2011) to frame the importance of productive classroomdiscourse for students' learning motivation. Interest is defined as aperson's repeated engagement or focused attention on an object,determined by a specific relationship between person and objectand shaped by interactions and the environment (Krapp, 2000;Renninger & Hidi, 2011). In self-determination theory (SDT) it isimportant that students perceive fulfilment of their basic psycho-logical needs during instruction. In this situation, they are morelikely to experience self-determined, intrinsic-learning motivation,and over time, these experiences are closely connected with apositive development of interest in the subject (Krapp & Prenzel,2011). Even though the potential for interest lies within the per-son, content and interaction define the development of situationaland individual interests (Hidi & Renninger, 2006). Thus, significantothers, the organisation of the environment and a person's innermotivational resources can support interest development(Renninger & Hidi, 2002; Sansone & Smith, 2000).
SDT proposes that a person's motivation depends on the fulfil-ment of three basic psychological needs: autonomy, competenceand social relatedness. The fulfilment of these needs depends on aperson's interpretation of the social contextdwhether the func-tional significance of a situation is controlling or informational(Ryan & Deci, 2002). Motivation is conceptualised as a situationalconstruct dependent on a person's moment-to-moment experi-ences and the interpretation. By changing students' moment-to-moment classroom experiences, e.g. through productive class-room discourse, their need for autonomy, competence and socialrelatedness is repeatedly fulfilled, leading to self-determined-learning motivation. Thus, teachers can create repeated instancesof triggered situational interest (Hidi & Renninger, 2006; Krapp,2005; Renninger & Hidi, 2011) using teaching strategies that pro-mote productive classroom discourse.
Grounded in the situational beneficial effects of productiveclassroom discourse on students' self-determined-learning moti-vation, we presume that prolonged positive experiences with basicpsychological need support and the resulting motivational out-comes (intrinsic learning motivation) will crystallise in the moreenduring motivational orientation of interest in the subject (Fig. 2).The TPD programme in this study aims at improving teachers' skillsof creating productive classroom discourse, which should lead topositive changes in self-determined-learning motivation throughrepeated instances of basic need fulfilment and thus to positivelydeveloping students' interest in the subject.
In conceptualising a study design that measures students'situational experiences at multiple points in time as well as theiroverall development over the course of a school year, we aim tocombine these two stances on students' learning motivation toshed more light on the importance of day-to-day classroom expe-riences and the learning environment for students' development ofself-determined-learning motivation and interest in the subject.Furthermore, we investigate whether TPD programmes can helpteachers counter students' decreases in learning motivation andinterest throughout secondary school (Turner, Warzon, &Christensen, 2011).
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e10396
1.1.3. The ‘Dialogic Video-Cycle’dan innovative TPD programmetargeting productive classroom discourse
Given the predominance of narrowly focused classroomdiscourse, supporting teachers to change their discourse behaviouris a palpable method of classroom reform. On the basis of theteaching cycle idea of planning, teaching and reflecting (Borko,Jacobs, Eiteljorg, & Pittman, 2008), Gr€oschner, Seidel, Kiemer, etal. (2014), Gr€oschner, Seidel, Pehmer, and Kiemer (2014) devel-oped a video-based TPD intervention for redefining classroomdiscourse practices (‘Dialogic Video-Cycle’, DVC) and showed howto implement it successfully with mathematics and science teach-ers (Gr€oschner, Seidel, Kiemer, et al., 2014).
The intervention group teachers participated in two iterations ofthe DVCda newly developed TPD programme adhering to effectivecomponents of TPD found in the literature (Desimone, 2009; vanVeen, Zwart, & Meirink, 2012; Wilson, 2013). The DVC concen-trates on generic aspects of classroom discourse as part of generalpedagogical knowledge. It implements productive classroomdiscourse in the form of two main activities: student activation andclarifying discourse rights and scaffolding student ideas and feedback(Walshaw & Anthony, 2008). The DVC aims to change teachers'perspective towards student learning processes. Each cycleincluded three workshops and one lesson videotaping (Fig. 1). Inthe first workshop, teachers received input on productive class-room discourse by a facilitator, and together with the facilitator andin collaborative practice, they adapted existing lesson plans bytaking concrete activities of productive classroom discourse (inaccord with Walshaw & Anthony, 2008) into account (Gr€oschner,Seidel, Kiemer, et al., 2014). Next, they were videotaped by theresearch team while teaching the revised lesson. The facilitatorchose video excerpts on the basis of the criteria of productiveclassroom discourse and prepared them as a basis for teacherreflection in workshops 2 and 3. Workshop 2 focused on activity 1(student activation), and teachers exchanged ideas about thediscursive roles of teachers and students and the ways in whichstudents are engaged. Workshop 3 focused on activity 2 (scaffoldingstudent ideas), and teachers exchanged ideas about how to take up
Fig. 1. Treatment conditions in DVC as intervention (left) a
student responses and elaborations (e.g. right/wrong answers, newideas, misconceptions etc.) and give feedback. During the work-shops, they watched selected clips, clarified questions about pro-ductive classroom discourse and jointly reflected on theirexperiences. The facilitator posed guiding questions (e.g. ‘Whichstrategies does the teacher in the video clip use to promote studentactivation?’); the group discussed what attracted their attentionand gave feedback (including solutions and alternatives), or askedmore questions. All discussions were chaired by the facilitator(Gr€oschner, Seidel, Pehmer, et al. 2014). The second iteration of theDVC followed the same course of action.
The DVC was implemented as a year-long TPD programme (twoiterations) in 10 STEM classrooms with the aim to improve teach-ers' classroom discourse practices and their students' perceptionsof motivation-enhancing learning environments. After ensuring thefidelity of implementation (Gr€oschner, Seidel, Kiemer, et al., 2014),in this paper, we explore the effects of the DVC on teachers'discourse practices (questioning and feedback behaviour) andstudents' self-determined learning motivation. The aim of thestudy is, therefore, to investigate in detail the DVC's effect on stu-dents' perception of basic psychological needs and self-determinedlearning motivation as well as the more enduring motivationalcharacteristic of interest in the subject, compared with that in thetraditional TPD format in Germany. Thus, we contribute to theresearch field by applying a longitudinal design, integrating situa-tional student perceptions of the learning environment with long-term developments and fostering dialogue in STEM classrooms bytaking into account both teachers' practice and students'perceptions.
1.2. Research questions
Four main questions were addressed:
(1) Do teachers' practices in the IG change towardsmore dialogicteaching (specifically questioning and feedback) in thecourse of the DVC, compared with teachers in a CG?
nd the traditional programme as control group (right).
Fig. 2. Relationship between basic psychological needs, intrinsic learning motivation and motivational student orientation over time.
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103 97
Hypothesis 1: Teachers in the IG will exhibit an increasednumber of open questions and constructive feedback (Hypothesis1a) and a decreased number of closed questions and simple feed-back in a preepost comparison (Hypothesis 1b), compared withthat for the CG teachers.
(2) Does a one-year DVC intervention show positive effects onstudents' development of interest in the subject (IG)compared with that for the CG?
Hypothesis 2: The students of teachers participating in the DVCwill show a positive development of domain-specific interest at theend of the school year compared with the students of teachers inthe CG.
(3) Are positive changes during the DVC intervention in the IGobserved regarding
a. students' perception of their teachers' support of theirbasic psychological needs?b. students' experience of self-determined-learning
motivation?
Hypothesis 3: The video-based intervention will lead to positivechanges in (a) students' perception of basic psychological needssupport and (b) their experience of self-determined-learningmotivation over the course of the year.
(4) Are the positive changes in interest in the subject at the endof the school year systematically related to changes in self-determined-learning motivation, which in turn is systemat-ically related to a change in teachers' support of students'basic psychological needs?
Hypothesis 4: Repeated altered experiences with regard to basicpsychological need perception (increased fulfilment of basic psy-chological needs) will lead to an increase in self-determinedlearning motivation. In turn, the repeated experience of self-determined learning motivation and situational interest willmanifest in an increased dispositional domain-specific interest.
2. Method
2.1. Participants
The participants were 10 teachers and their 226 students (47.8%girls, 52.2% boys) in 10 science and math classrooms from Germanmiddle- or high-tracked schools. The students' mean age was 15.67years (SD ¼ .98); they were in 9th grade.
In the context of German TPD conditions, the teachers chosewhether they wanted to participate in the video-based TPD pro-gramme (intervention group, IG) or programme with a set ofworkshops (control group, CG). They did not know which groupserved as the intervention and which was the control condition.This avoided having the teachers feel reluctant to participate in thevideo-based TPD programme or doubt the effectiveness of watch-ing themselves (Fishman, Marx, Best, & Tal, 2003). Six teachersopted for the intervention (DVC), and four chose the traditional TPDprogramme. The two groups did not differ in age (M ¼ 38.3,SD ¼ 5.56, U(6:4) ¼ 7.5, z ¼ �.97, p ¼ .33), teaching experience(M ¼ 5.65, SD ¼ 2.93, U(6:4) ¼ 6.0, z ¼ �1.30, p ¼ .20) or gender,c2gender(df ¼ 1) ¼ .08, p ¼ .79. Both groups showed comparativeengagement (and time spent) in prior TPD programmes. They didnot differ in their motivation (‘I want to learn more about commu-nication in the classroom’,MIG¼ 3.51, SD¼ .47;MCG¼ 3.81, SD¼ .38;U(6:4) ¼ 7.0, z ¼ �1.14, p ¼ .25) for the respective TPD programme
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e10398
at the outset of the study or any of the other considered backgroundvariables such as basic need fulfilment and satisfaction with theprogramme (Gr€oschner, Seidel, Kiemer, et al., 2014). Video as a toolfor reflection was only used in the IG.
Given their teachers' choices, there were 136 students in the IGand 90 students in the CG. The two groups differed regarding age,t(224) ¼ 5.20, p < .01, d ¼ .71 (IG: Mage ¼ 15.41, SD ¼ .98; CG:Mage ¼ 16.07, SD ¼ .85) and gender, c2gender(df ¼ 1) ¼ 8.94, p < .01(IG: 39.7% girls; CG: 60.0% girls), for which reason both variableswere accounted for as covariates during the data analysis.
2.2. Research design
Data were collected in a one-year longitudinal two-group-intervention design. Fidelity of implementation was tested andproven successful (Gr€oschner, Seidel, Kiemer, et al., 2014). The studyran during the school year 2011/2012. The two treatments includedthe same content (classroom discourse) and had the same duration(22 h). They differed with regard to relating TPD content to actualteaching practice by providing opportunities to reflect on changedpractice in a community of learners.
Students in both groups were questioned regarding motiva-tional orientations at the beginning (pre) and end (post) of thestudy. The IG students were also questioned directly after video-taping one lesson used for reflective purposes in the DVC (DVC 1,DVC 2). Videotaping took place at the beginning and at the end ofthe school year in both groups, and additionally, twice as part of theintervention (DVC 1, DVC 2) for the IG.
Dialogic Video-Cycle (IG): The IG teachers participated in a TPDprogramme with two iterations of the DVC (see Section 3.1). Eachcycle included three workshops and one lesson videotaping (Fig. 1).In workshop 1, teachers focussed on revising a lesson plan byconsidering productive classroom dialogue elements and commu-nication strategies. Workshop 2 focused on student activation,mainly targeted through teachers' questioning behaviour. Scaf-folding of students ideas, the focus of workshop 3, was targetedespecially through teacher feedback.
Advanced Traditional Programme (CG): The CG teachers chose aset of workshops with a focus on productive classroom discourseoffered by the local TPD institute (Fig. 1). They met twice inroundtables (hosted by the same facilitator as in the IG) to sharetheir TPD experiences. They were encouraged to share how theyhad experienced the central aspects of productive classroomdiscourse during their TPD and how they would apply their expe-rience to teaching.
2.3. Instruments
All items were answered on a 4-point-Likert scale ranging from0 ¼ ‘strongly disagree’ to 3 ¼ ‘fully agree’.
2.3.1. QuestionnairesThe scale for interest in the subject was used prior in the
German context of PISA (Ramm et al., 2006) to assess situationalperceptions in a national video study (Seidel, Prenzel, Duit, &Lehrke, 2003). Interest in the subject was assessed on the basis ofa pre- and post-test. The four scales of perception of basic psy-chological need fulfilment and self-determined learningmotivationwere administered after the four videotaped lessons: pre-test,during DVC 1, during DVC 2 and post-test.
Interest in the subject: This scale consisted of five items (‘I aminterested in mathematics/science’) and showed good reliability(a ¼ .84, a ¼ .87).
Basic psychological need fulfilment: Autonomy perceptions wereassessed with eight items (‘During this class, I had the feeling that the
teacher was open for different student answers’), competence per-ceptions with seven items (‘During this class, I had the feeling thatthe teacher thought us capable of challenging exercises’) and socialrelatedness with six items (‘During this class, I felt like I wasimportant to the teacher’). All scales showed satisfactory to goodreliability scores: autonomy a ¼ .75e.82, competence a ¼ .71e.83and social relatedness a ¼ .69e.85.
Intrinsic learning motivation: This scale consisted of three items(‘This classwas fun’)with reliability scores ranging froma¼ .75 to .85.
2.4. Video analysis
Using the software Videograph (Rimmele, 2002), we analysedteacherestudent interactions focussing on their quality of class-room discourse, specifically the facilitation of activities 1 (studentactivation, characterised by questioning behaviour) and 2 (scaffoldingof student ideas, characterised by feedback behaviour) on the basis ofthe method by Walshaw and Anthony (2008). The video analysisemployed event sampling, using speaker turns (teacher, student),as the unit of analysis (k ¼ 1.0). A coding system was adapted anddeveloped by the research team (Pehmer, Kiemer, & Gr€oschner,2014). All categories focused on discourse behaviour duringwhole class discussions (teacher questions and feedback). Fivemonths of training preceded independent raters' scoring of thevideos:
Teacher questioning: Independent raters classified each questionas open (‘What do you think happens if we heat it up?’) or closed (‘Dowe have any right angles here?’). Inter-rater reliability in this cate-gory was k ¼ .79 (direct agreement: 89.7%).
Teacher feedback: Teachers' feedback was characterised asconstructive (‘That's a good strategy; just try to focus more on themechanism’) or simple (‘Nice job’). The reliability between coderswas k¼ .71, and direct agreement (85.3%) was satisfactory (Landis&Koch, 1977).
2.5. Data analysis
Research Question 1: To answer the first research question, weperformed a repeated measure non-parametric analysis of variance(Brunner, Domhof, & Langer, 2002) for the pre- and post-testcomparisons, using treatment as the independent variable (IV)and behavioural indicators of teachers' productive classroomdiscourse as dependent variables (DV).
Research Question 2: A latent growth model, using treatment aspredictor variable (dummy coded: IG ¼ 1, CG ¼ �1) over twomeasurement points (pre-to post-test) was created to account forthe nested structure of the data (Duncan & Duncan, 2004; Geiser,2011). Since only two measurement points (pre- and post-test forIG and CG) were taken into consideration, only a linear growthcurve model could be applied. A test of covariance structureshowed an autoregressive covariance structure of the first order tofit the data best (c2(df¼ 1)¼�1.33; p < .01). Age and sex were usedas covariates, given the a priori differences in group composition(Shek & Ma, 2011).
Research Question 3: Given the nested data structure, growthcurve models (linear, quadratic and cubic) over four measurementpoints within the IG were run again, both for support of basicpsychological needs (perceived autonomy, competence and socialrelatedness) and intrinsic learning motivation. The informationcriteria (�2 log likelihood, AIC, BIC) in each case were smallest forthe cubic models, and the c2-ratio tests between nested modelsshowed that the cubic model provided a better fit than the moreparsimonious counterparts for all variables. The best-fit covariancestructure was provided by a first-order autoregressive structure(Shek & Ma, 2011).
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103 99
Research Question 4: Bivariate correlational analyses betweenthe delta values from pre- to post-test of all basic psychologicalneeds and the delta value of intrinsic learningmotivation, as well asinterest in the subject, were run. Additionally, we conducted a pathanalysis modelling the predictive power of the change in perceivedautonomy and competence for the change in intrinsic learningmotivation and interest in the subject (Geiser, 2011).
3. Results
3.1. Teachers' changed practices
All reported results are relative counts of the respective sub-category in relation to the total in that category. Table 1 gives thedescriptive statistics and non-parametric analyses of variance ofthe video analysis of teachers' questioning behaviour (activity 1)and feedback (activity 2) for both groups at both measurementpoints. For the IG teachers' questioning behaviour, the number ofopen questions increased and the number of closed questionsdecreased; yet this trend did not reach significance (Fopen(1) ¼ .56,n.s.; Fclosed(1) ¼ .56, n.s.). For their feedback behaviour, there was asignificant increase in constructive feedback (F(1) ¼ 9.20, p < .01)and a significant decrease in simple feedback (F(1) ¼ 9.36, p < .01).
3.2. Effect on students' interest in the subject
A preliminary inspection of mean values (MIG ¼ 1.48, SD ¼ .67;MCG ¼ 1.33, SD ¼ .66) showed comparable scores between bothgroups at pre-test. Descriptive comparisons between the meanscores of the pre- and post-test show a slight increase in interest inthe subject in the IG (Mpost¼ 1.55, SD¼ .75) but a decrease in the CG(Mpost ¼ 1.25, SD ¼ .67).
The unconditional linear growth model exhibited a significantintercept (b ¼ 10.76, SE ¼ 1.89, p < .01) suggesting that the level ofinterest was not constant over time, while the non-significant slopesuggests a constant positive growth rate in the level of interest inthe subject over time across both groups (b ¼ 3.04, SE ¼ 2.70, n.s.).The inter-class correlation for this model (ICC ¼ .23) showed that23% of the variation in interest scores is attributable to the inter-individual level. To account for group differences, the factor treat-ment was introduced into the model as a predictor variable. Whiletreatment was not associated with the initial level of interest in thesubjects (b¼�2.64, SE¼ 2.07, p¼ n.s.), it is a significant predictor ofthe linear growth rate in interest between the two groups (b¼ 5.87,SE¼ 2.96, p¼ .05). There was a differential positive development ofinterest in the subject in favour of the IG. A change in residualvariance of 4.05 between the two models attributes 4.05% of thevariation in interest scores to group membership.
Table 1Descriptive statistics of teachers' questions and feedback.
Pre-test Post-test
M SD Meanrank
RTE M SD Meanrank
RTE
Closed questions IG .59 .21 8.08 .38 .61 .18 12.13 .58CG .75 .14 8.33 .39 .81 .07 15.75 .76
Open questions IG .40 .21 12.91 .62 .39 .17 12.67 .61CG .23 .17 8.88 .42 .17 .09 5.25 .24
Simple feedback IG .79 .07 13.00 .63 .58 .14 5.08 .23CG .71 .11 9.00 .43 .85 .10 16.38 .79
Constructive feedback IG .21 .07 8.00 .38 .39 .14 15.92 .77CG .29 .11 12.00 .58 .15 .10 4.63 .21
Note. n(IG) ¼ 6 teachers, n(CG) ¼ 4 teachers; M ¼ mean, SD ¼ standard deviation;values are relative percentages; RTE ¼ relative treatment effect.
3.3. Effects on changes in basic psychological need perceptions andself-determined-learning motivation in the DVC
In answer to research questions (3a) and (3b), we focussed onthe IG and investigated trajectories in students' perceptions of theirbasic psychological need fulfilment and intrinsic learning motiva-tion over fourmeasurement points. Table 2 provides the descriptivestatistics for these variables for all measurement points.
An inspection of the mean values of the variables across timesuggests curvilinear trajectories. In all cases, curvilinear models fitthe data best.
Autonomy perceptions: The cubic growth curve model showsbetter information criteria than its more parsimonious counter-parts (�2 log likelihood ¼ 697.88, AIC ¼ 711.88, BIC ¼ 741.03). A c2-ratio test supported the best fit to the data for this model(c2(df¼ 2)¼�25.47). Themodel intercept was significant (b¼ 1.72,SE¼ .05, p < .01), as was the estimate of the cubic growth (b¼ 2.39,SE ¼ .67, p ¼ .01). The ICC was .33.
Competence perceptions: As in the case of autonomy, the curvi-linear growth model for competence had the best values in theinformation criteria (�2 log likelihood ¼ 638.29, AIC ¼ 652.29,BIC ¼ 681.43). A c2-ratio test showed that it is superior to thesimpler linear (c2(df ¼ 2) ¼ �15.19) growth model. The modelexhibited a positive significant estimate for intercept (b ¼ 1.98,SE¼ .04, p< .01) and the cubic growth parameter (b¼ 2.13, SE¼ .58,p < .01). The ICC of the model was .42.
Social relatedness perceptions: For social relatedness, no signifi-cant growth trajectory from pre-to post-test could be found, yet thecubic model showed the best overall model fit (�2 loglikelihood ¼ 728.67, AIC ¼ 742.67, BIC ¼ 771.81, c2(df ¼ 2) ¼ �5.29).It revealed a significantly positive linear growth between pre-testand DVC 1 (b ¼ .83, SE ¼ .42, p ¼ .05). The non-significant trendsfor a negative quadratic growth (b¼�1.77, SE¼ 1.03, p¼ .09) and apositive cubic growth (indicated by a b-estimate of .92; p ¼ .17)showed that after increasing significantly from pre-test to DVC 1,students' perceptions of teachers' social relatedness supportremained levelled. The ICC for this model was .36.
Intrinsic learning motivation: A cubic growth curve model pro-vided the best fit to the data (�2 log likelihood ¼ 974.19,AIC ¼ 988.19, BIC ¼ 1017.39, c2(df ¼ 2) ¼ �11.34), producing asignificant intercept (b ¼ 1.66, SE ¼ .06, p < .01) and cubic slopeestimate (b ¼ 1.75, SE ¼ .83, p ¼ .04). The ICC of the model was .55.
3.4. Student interest and its relation to changes in self-determinedintrinsic learning motivation in the DVC
First, bivariate correlations for Dpreepost scores of basic psy-chological need support and intrinsic learning motivation, as wellas interest in the subject were run for the IG. As Table 3 illustrates,the correlational analyses revealed systematic positive relation-ships between the changes from pre- to post-test for all situationalmotivational variables and students' interest in the subject, exceptfor social relatedness. Only autonomy and competence were used
Table 2Descriptive statistics for basic psychological needs and intrinsic learning motivationin the IG.
Pre-test DVC 1 DVC 2 Post-test
M SD M SD M SD M SD
Autonomy 1.74 .55 2.06 .52 1.88 .54 1.90 .57Competence 2.02 .48 2.18 .52 2.02 .57 2.10 .59Social relatedness 1.92 .53 2.04 .60 1.95 .51 1.91 .61Intrinsic learning motivation 1.67 .73 1.91 .73 1.73 .74 1.71 .74
Note. n ¼ 136 Students; 0 ¼ ‘strongly disagree’ to 3 ¼ ‘fully agree’.
Fig. 3. Path model of changed basic psychological need fulfilment, intrinsic learningmotivation and interest in the subject.
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103100
in the path analysis because social relatedness did not show sig-nificant preepost changes in the growth curve models nor did itshow significant bivariate correlations.
Overall, the model fits the data very well (c2(df ¼ 2) ¼ .82,p ¼ .66; RMSEA ¼ .00, CFI ¼ 1.0). As expected, the change in situ-ational intrinsic learning motivation is significantly predicted bychanges in perceived autonomy (b ¼ .27**, SE ¼ .09) and experi-ences of competence (b ¼ .31**, SE ¼ .08). Further, the change insituational intrinsic learning motivation is then predictive of thechange in interest in the subject over the course of the intervention(b ¼ .48**, SE ¼ .08). Changes in autonomy and competence expe-riences also have an additional indirect effect on the change ininterest via intrinsic learningmotivation (bauton_ind¼ .13**, SE¼ .05;bcomp_ind ¼ .12**, SE ¼ .04). Fig. 3 depicts the model with b-weights.The model explained 21% of the changes in intrinsic learningmotivation (R2 ¼ .21, p < .01) and 23% in interest in the subject(R2 ¼ .23, p < .01).
4. Discussion
This study investigated the extent to which a video-based TPDintervention targeting productive classroom discourse showedpositive effects on teachers' practice and students' motivation tolearn mathematics and science. Thus far, the majority of studiestargeting productive classroom discourse have focussed on the co-construction of knowledge and achievement as outcome variables(Lipowsky et al., 2007) or qualitative analysis of classroom in-teractions (Michaels&O'Conner, 2012). Thus, this to our knowledgeis the first study focussing explicitly on teachers' changed practicesand students' motivation to learn mathematics and science. Itfurthermore takes both a situational and longitudinal stance onstudents' self-determined learning motivation and brings togetherstate of the art research on productive classroom discourse(Walshaw & Anthony, 2008; Wells & Arauz, 2006) and the role ofteachers and TPD in supporting students' self-determined learningmotivation (Turner et al., 2011). After describing the successfulimplementation of a TPD programme focussing on productiveclassroom discourse (Gr€oschner, Seidel, Kiemer, et al., 2014), weshowed that the reported decrease in students' self-determinedlearning motivation and interest (Eccles et al., 1993; Maulanaet al., under review) can be countered to a certain extent byteachers' facilitation of productive classroom discourse. Hereby, theDVC took the components of effective TPD into account (Gr€oschner,Seidel, Kiemer, et al., 2014; Desimone, 2009; van Veen et al., 2012;Wilson, 2013) and included the use of video in the cycle of planning,teaching and reflecting (Santagata, 2009). The intervention wascompared to a traditional TPD programme including a coherent setof workshops on the same content, with the same duration, and anadditional opportunity for social exchange among participants.
The video analysis of teachers' discourse behaviour at pre- andpost-test of the study indicated that the IG teachers changed theirquestioning and feedback behaviour positively (Hypothesis 1a and
Table 3Changes from pre-to post-test in motivational constructs: bivariate correlations.
Interest inthe subjectDpreepost
Intrinsic learningmotivationDpreepost
AutonomyDpreepost
CompetenceDpreepost
Social relatednessDpreepost
1 1 .28** .25** .36** .17þ
2 1 .47** .38** .25**3 1 .52** .46**4 1 .48**5 1
Note. n ¼ 108 Students; þ<.10, *<.05, **<.01.
1b confirmed). While the CG teachers showed increasingly moreclosed questions and simple feedback, indicating a more teacher-centred and less productive discourse behaviour, participants inthe video-based, reflection-oriented treatment condition (DVC)exhibited a positive trend and significant positive changes in thecourse of the study. The significant increase in constructive feed-back shows that this aspect of productive classroom discourse isapparently more likely to be demonstrated in observed practicesthan amore open questioning style (van den Bergh, Ros,& Beijaard,2013a, 2013b). Thus, questioning seems to be a more-difficult-to-adapt teacher practice as teachers' style of asking questions isvery much a routine (Oliveira, 2010) and teachers sometimesstruggle with asking open questions because they require differentsituational interactions with students, whereas feedback is more ofan instructionmatter (Mercer, 2010). Furthermore, even though theCG received a form of TPD, the lack of opportunities for video-basedreflection, using examples from their own practice and a strongcommunity of learners (McLaughlin & Burnaford, 2007) may havehindered their transfer of new practices into their classrooms andthus accounts for their decrease in productive classroom discoursepractices.
With regard to the second research question, this study showedthat compared with a CG, the DVC resulted in more positive lineargrowth of students' interest in the subject. These kinds of decreasesare also observed in the CG, but not for teachers participating in theDVC (Hypothesis 2 confirmed). Thus, teacher participation in theDVC intervention on re-defining classroom discourse is an appro-priate means to improve student interest in STEM subjects. A b-weight of 5.87 on this motivational student characteristic un-derscores the effectiveness of the DVC as a relevant, practice-oriented TPD programme (Gr€oschner, Seidel, Kiemer, et al., 2014).Furthermore, the positive comparison of the DVC with the CGsuggests that the TPD programme may be a more powerful andeffective form of professional development (Borko et al., 2008;Sherin & Han, 2004) than traditional workshops.
Furthermore, this finding adds to the body of literature thatconceptualises interest as a state-trait-like variable that can bedeveloped through repeated positive interactions between personand object, as well as by the specific characteristics of the envi-ronment (Frenzel, Pekrun, Dicke, & Goetz, 2012; Sansone & Smith,2000). The findings support those theories of interest developmentthat propose situational interest as an antecedent of sustained in-dividual interest and thus as a prerequisite for interest
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103 101
development (Hidi & Renninger, 2006; Krapp & Prenzel, 2011).Also, we present further evidence that productive classroomdiscourse positively affects motivational student variables (Juriket al., 2014).
The DVC, by focussing on teachers' use of strategies for pro-ductive classroom discourse (Walshaw & Anthony, 2008), wasconceptualised in a way that students' need for autonomy,competence and social relatedness was scaffolded. In line with theliterature on both self-determination theory (Ryan & Deci, 2000)and productive classroom discourse, we assumed that these kind ofchanged experiences are connected to an increase in self-determined-learning motivation. The study results support theseconjectures and accord with previous findings of the effects ofproductive classroom discourse on student learning outcomes(Kovolainen & Kumpulainen, 2005; Walshaw & Anthony, 2008).Hypothesis 3a was therefore confirmed. The curvilinear trajectoriesof the IG teachers for autonomy and competence support are in linewith the literature on teacher learning. Existing research oftenwarrants fluctuation in such learning because of teachers grapplingwith new practices and the time needed to consolidate changes inroutine practices (Bakkenes, Vermunt, & Wubbels, 2010), and itsupports the often cited argument in TPD research for continuous,long-term teacher learning (Borko et al., 2008). In the case of socialrelatedness, fluctuations were less strong, resulting in a significantinitial increase, which then levels out throughout the school year. Apossible explanation for this different trend could be that theclasses had not been newly formed at the beginning of the schoolyear, but as is the norm in Germany, had been a group (with fewchanges) since 5th grade. Most classes had the same teacher inprevious school years; hence, it can be assumed that a feeling ofconnectedness had already been established and was onlyimproved slightly through the intervention (Mainhard,Brekelmans, & Wubbels, 2011). Future research should focus onthis issue by inquiring into the specific mechanisms of productiveclassroom discourse that promote basic psychological needperception, and in turn, self-determined-learning motivation bytaking class transition processes or peer group-specific learningprocesses into account.
Regarding our fourth research question, the bivariate correla-tions in changed student perceptions of basic psychological needsupport, intrinsic learning motivation and interest in the subject,indicate that the positive differences in these variables are signifi-cantly related. The path analyses support that continuous changesin situational perceptions of autonomy and competence are pre-dictive of students' changed perceptions of intrinsic learningmotivation. Beyond that, perceptions of autonomy and competenceare also indirectly predictive of students' experiences of interest inSTEM subjects, as is their perception of their situational intrinsiclearning motivation. These findings are in line with reported effectsof productive classroom discourse on student motivation (Juriket al., 2014) and research on the development of interest in thesubject (Frenzel et al., 2012; Renninger & Hidi, 2011). Hypothesis 4could thus be confirmed. These instances of perceived autonomyand competence can be enhanced through teaching practices thatfocus on productive classroom discourse, or more specifically, onstudent activation and scaffolding of student ideas (Walshaw &Anthony, 2008). The next steps are to further investigate the rela-tion between perceived autonomy/competence and classroomdiscourse, as well as the specific role of the teachers' support forbasic need fulfilment in the context of classroom discourse.Furthermore, aspects of teachers' attitudes and beliefs as pre-requisites for changed practice should be considered.
In interpreting the study results, the small and non-randomisedsample has to be taken into account. A year-long intervention study,especially one using classroom videos is an invasive method that a
small number of teachers are willing to undergo (van Eekelen,Vermunt, & Boshuizen, 2006). Yet this holds both for the DVCintervention teachers as well as the CG since all teachers had toengage with the same effort with respect to time and content. Toguarantee the participants' acceptance and minimise reactance,they were given a free choice between the two TPD programmes,not randomly assigned. As this was the first study exploring theeffects of the DVC, we focussed on not compromising the quality ofthe data by reactive behaviours during the school year (Gr€oschner,Seidel, Kiemer, et al., 2014). Future studies should increase theteacher sample to avoid a selection bias and allow more general-isable inductions from the data, as well as apply hierarchical latentgrowth curve models to consider the nested data-structure at theteacher level. Another limitation of this study is its reliance onstudents' self-reports to establish their perceptions of basic psy-chological need fulfilment and their motivational orientations.Further investigations should attempt to diversify data on students'perceptions, e.g. through the use of teacher reports, behaviouralindicators or techniques like experience sampling.
Future research should more specifically investigate whichteacher practices foster students' perceptions of teachers' basicneed support and the role of students' basic need experiences.Given the importance of both perceived autonomy and competenceas predictor for changes in students' interest in the subject, in-vestigations into ways of fostering perceptions of autonomy/competence, as well as teachers' autonomy-support/competence-support through productive classroom discourse appear espe-cially fruitful (Reeve & Jang, 2006). More fine-grained analyses ofthe micro-processes of teacherestudent interactions could shedlight on differential effects of teacher discourse behaviour (feed-back and questioning) on students' basic need perceptions. Mixed-method approaches that combine student self-report and obser-vational data appear especially meaningful in this context.
5. Conclusions
This study shows that after successful implementation(Gr€oschner, Seidel, Kiemer, et al., 2014), the video-based TPDapproach of the DVC was effective in changing teachers' behaviourtowards more productive classroom discourse. This finding isespecially noteworthy because the comparison with the CG showsthat without such effective measures, teachers tend to narrow theirdiscourse practices (in the form of closed questions and simplefeedback) towards more teacher-centred forms of discourse overthe course of a school year. The results of this study further showpositive changes in students' experiences of autonomy, compe-tence and social relatedness as well as intrinsic learningmotivation,when their teachers participated in the DVC intervention. This isthe first study to integrate student motivation research withresearch on productive classroom discourse supporting the theo-retically derived assumptions about changes in students' motiva-tion to learn mathematics and science because of specific TPDactivities of their teachers. The results demonstrate the importanceof productive classroom discourse in promoting positive learningoutcomes for students' motivational orientations and its role infostering student interest in STEM subjects.
Acknowledgements
We would like to thank the teachers who participated in theproject “DIALOGUE”. This work was supported by the GermanResearch Foundation under Grant SE 1397/5-1.
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103102
References
Ainley, M., Hidi, S., & Berndorff, D. (2002). Interest, learning, and the psychologicalprocesses that mediate their relationship. Journal of Educational Psychology,94(3), 545e561. http://dx.doi.org/10.1037//0022-0663.94.3.545.
Bakkenes, I., Vermunt, J. D., & Wubbels, T. (2010). Teacher learning in the context ofeducational innovation: learning activities and learning outcomes of experi-enced teachers. Learning and Instruction, 20(6), 533e548.
van den Bergh, L., Ros, A., & Beijaard, D. (2013a). Feedback during active learning:elementary school teachers' beliefs and perceived problems. EducationalStudies, 39(4), 418e430. http://dx.doi.org/10.1080/03055698.2013.767188.
van den Bergh, L., Ros, A., & Beijaard, D. (2013b). Teacher feedback during activelearning: current practices in primary schools. British Journal of EducationalPsychology, 83(2), 341e362. http://dx.doi.org/10.1111/j.2044-8279.2012.02073.x.
Borko, H., Jacobs, J. K., Eiteljorg, E., & Pittman, M. E. (2008). Video as a tool forfostering productive discussions in mathematics professional development.Teaching and Teacher Education, 24(2), 417e436. http://dx.doi.org/10.1016/j.tate.2006.11.012.
Brunner, E., Domhof, S., & Langer, F. (2002). Nonparametric analysis of longitudinaldata in factorial experiments. Michigan: Wiley.
Deci, E., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in humanbehaviour. New York: Plenum.
Deci, E., & Ryan, R. M. (2004). In Handbook of self-determination research (Softcovered.). Rochester, NY: Univ. of Rochester Press.
Desimone, L. M. (2009). Improving impact studies of teachers' professional devel-opment: toward better conceptualizations and measures. EducationalResearcher, 38(3), 181e199. http://dx.doi.org/10.3102/0013189x08331140.
Duncan, T. E., & Duncan, S. C. (2004). An introduction to latent growth curvemodelling. Behavior Therapy, 35, 333e363.
Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Reuman, D., Flanagan, C., et al.(1993). Development during adolescence: the impact of stageeenvironment fiton young adolescents' experiences in schools and families. American Psycholo-gist, 48(2), 90e101. http://dx.doi.org/10.1037/0003-066X.48.2.90.
van Eekelen, I. M., Vermunt, J. D., & Boshuizen, H. P. A. (2006). Exploring teachers'will to learn. Teaching & Teacher Education, 22, 408e423. http://dx.doi.org/10.1016/j.tate.2005.12.001.
Fishman, B. J., Marx, R. W., Best, S., & Tal, R. T. (2003). Linking teacher and studentlearning to improve professional development in systematic reform. Teachingand Teacher Education, 19, 643e658. http://dx.doi.org/10.1016/S0742-051X(03)00059-3.
Frenzel, A. C., Pekrun, R., Dicke, A.-L., & Goetz, T. (2012). Beyond quantitativedecline: conceptual shifts in adolescents' development of interest in mathe-matics. Developmental Psychology, 48, 1069e1082. http://dx.doi.org/10.1037/a0026895.
Furtak, E. M., Seidel, T., Iverson, H., & Briggs, D. C. (2012). Experimental and quasi-experimental studies of inquiry-based science teaching: a meta-analysis. Re-view of Educational Research, 82(3), 300e329. http://dx.doi.org/10.3102/0034654312457206.
Geiser, C. (2011). Datenanalyse mit Mplus: Eine anwendungsorientierte Einführung.Wiesbaden: VS Verlag für Sozialwissenschaften.
Gr€oschner, A., Seidel, T., Kiemer, K., & Pehmer, A.-K. (2014). Through the lens ofteacher professional development components: the “dialogic video cycle” as aninnovative program to foster classroom dialogue. Professional Development inEducation. Online, http://dx.doi.org/10.1080/19415257.2014.939692.
Gr€oschner, A., Seidel, T., Pehmer, A.-K., & Kiemer, K. (2014). Facilitating collaborativeteacher learning: the role of “mindfulness” in video-based teacher professionaldevelopment programs. Gruppendynamik und Organisationsberatung, 45(3),273e290. http://dx.doi.org/10.1007/s11612-014-0248-0.
Harackiewicz, J. M., Durik, A. M., Barron, K. E., & Linnenbrink, E. A. (2008). The roleof achievement goals in the development of interest: reciprocal relations be-tween achievement goals, interest and performance. Journal of EducationalPsychology, 100, 105e122.
Hattie, J., & Timperley, H. (2007). The power of feedback. Review of EducationalResearch, 77(1), 81e112. http://dx.doi.org/10.3102/003465430298487.
Hidi, S., & Renninger, K. A. (2006). The four-phase model of interest development.Educational Psychologist, 41(2), 111e127. http://dx.doi.org/10.1207/s15326985ep4102_4.
Jurik, V., Gr€oschner, A., & Seidel, T. (2013). How student characteristics affect girls'and boys' verbal engagement in physics instruction. Learning and Instruction, 23,33e42. http://dx.doi.org/10.1016/j.learninstruc.2012.09.002.
Jurik, V., Gr€oschner, A., & Seidel, T. (2014). Predicting students' cognitive learningactivity and intrinsic learning motivation: how powerful are teacher state-ments, student profiles, and gender? Learning and Individual Differences,132e139. http://dx.doi.org/10.1016/j.lindif.2014.01.005.
Kovolainen, M., & Kumpulainen, K. (2005). The discursive practice of participationin an elementary classroom community. Instructional Science, 33, 213e250.http://dx.doi.org/10.1007/s11251-005-2810-1.
Krapp, A. (2000). Interest and human development during adolescence: an edu-cationalepsychological approach. In H. Heckhausen (Ed.), Motivational psy-chology of human development (pp. 109e128). London.
Krapp, A. (2002). An educational-psychological theory of interest and its relation toself-determination theory. In E. L. Deci, & R. M. Ryan (Eds.), The handbook of self-determination research (pp. 405e427). Rochester: University of Rochester Press.
Krapp, A. (2005). Basic needs and the development of interest and intrinsic moti-vational orientations. Learning and Instruction, 15, 381e395. http://dx.doi.org/10.1016/j.learninstruc.2005.07.007.
Krapp, A., & Prenzel, M. (2011). Research on interest in science: theories, methods,and findings. International Journal of Science Education, 33(1), 27e50. http://dx.doi.org/10.1080/09500693.2011.518645.
Krapp, A., & Ryan, R. M. (2002). Selbstwirksamkeit und Lernmotivation. Zeitschriftfür P€adagogik. Selbstwirksamkeit und Motivationsprozesse in Bildungsin-stitutionen, 44, 54e82.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement forcategorical data. Biometrics, 33(1), 70e78.
Lipowsky, F., Rakoczy, K., Pauli, C., Reusser, K., & Klieme, E. (2007). GleicherUnterricht e gleiche Chancen für alle? Die Verteilung von Schülerbeitr€agen imKlassenunterricht. Unterrichtswissenschaft, 35(2), 125e147.
Mainhard, T. M., Brekelmans, M., & Wubbels, T. (2011). Coercive and supportiveteacher behaviour: within- and across-lesson associations with the classroomsocial climate. Learning and Instruction, 21(3), 345e354. http://dx.doi.org/10.1016/j.learninstruc.2010.03.003.
Maulana, R., Opdenakker, M.-C., & den Brok, P. (2013). Teachers' instructional be-haviours as important predictors of academic motivation in middle school earlyadolescents: Changes and links across the school year (under review).
McLaughlin, H. J., & Burnaford, G. E. (2007). Re-thinking the basis for ‘high quality’teaching: teacher preparation in communities. In T. Townsend, & R. Bates (Eds.),Handbook of teacher education (pp. 331e342). Netherlands: Springer.
Mercer, N. (2010). The analysis of classroom talk: methods and methodologies.British Journal of Educational Psychology, 80(1), 1e14. http://dx.doi.org/10.1348/000709909x479853.
Michaels, S., & O'Conner, C. (2012). Talk science primer. Cambridge, MA: TERC.Oliveira, A. W. (2010). Improving teacher questioning in science inquiry discussions
through professional development. Journal of Research in Science Teaching,47(4), 422e453. http://dx.doi.org/10.1002/tea.20345.
Organisation for Economic Co-operation and Development. (2007). PISA 2006:Science competencies for tomorrow's world. Paris: OECD.
Pehmer, A.-K., Kiemer, K., & Gr€oschner, A. (2014). Produktive Lehrer-Schüler-Kom-munikation: Ein Kategoriensystem zur Erfassung Produktiver Gespr€achsführung imKlassengespr€ach und in Schülerarbeitsphasen. München: TUM School of Educa-tion, TUM.
Ramm, G., Prenzel, M., Baumert, J., Blum, W., Lehmann, R., Leutner, D., et al. (2006).PISA 2003. Dokumentation der Erhebungsinstrumente. Münster: Waxmann.
Reeve, J. M., & Jang, H. (2006). What teachers say and do to support students' au-tonomy during a learning activity. Journal of Educational Psychology, 98(1),209e218. http://dx.doi.org/10.1037/0022-0663.98.1.129.
Renninger, K. A., & Hidi, A. (2002). Student interest and achievement: develop-mental issues raised by a case study. In A. Wigfield, & A. J. Eccles (Eds.),Development of achievement motivation (pp. 173e195). New York, NY: AcademicPress.
Renninger, K. A., & Hidi, S. (2011). Revisiting the conceptualization, measurement,and generation of interest. Educational Psychologist, 46(3), 168e184. http://dx.doi.org/10.1080/00461520.2011.587723.
Renninger, K. A., Hidi, S., & Krapp, A. (1992). In The role of interest in learning anddevelopment. Hillsdale, NJ: Lawrence Erlbaum Associates.
Richter, D., Kunter, M., Klusmann, U., Lüdtke, O., & Baumert, J. (2011). Professionaldevelopment across the teaching career: teachers' uptake of formal andinformal learning opportunities. Teaching & Teacher Education, 27(1), 116e126.
Rimmele, R. (2002). Videograph. Kiel: IPN.Roth, K. J., Druker, S. L., Garnier, H. E., Lemmens, M., Chen, C. W., Kawanaka, T., et al.
(2006). Teaching science in five countries: Results from the TIMSS 1999 video studystatistical analysis report. Washington, DC: National Center for EducationStatistics.
Ryan, R. M., & Deci, E. (2000). Intrinsic and extrinsic motivations: classic definitionsand new directions. Contemporary Educational Psychology, 25(1), 54e67. http://dx.doi.org/10.1006/ceps.1999.1020.
Ryan, R. M., & Deci, E. (2002). Overview of self-determination theory: an organismicdialectic perspective. In E. Deci, & R. M. Ryan (Eds.), Handbook of self-determination research (pp. 3e33). Rochester, NY: The University of RochesterPress.
Sansone, C., & Smith, J. L. (2000). Interest and self-regulation: the relation betweenhaving to and wanting to. In C. Sansone, & J. Harackiewicz (Eds.), Intrinsic andextrinsic motivation: The search for optimal motivation and performance (pp.342e371). New York, NY: Academic Press.
Santagata, R. (2009). Designing video-based professional development for mathe-matics teachers in low-performing schools. Journal of Teacher Education, 60(1),38e51. http://dx.doi.org/10.1177/0022487108328485.
Santagata, R., & Guarino, J. (2010). Using video to teach future teachers to learn fromteaching. ZDM: The International Journal on Mathematics Education. http://dx.doi.org/10.1007/s11858-010-0292-3.
Schiefele, U., Krapp, A., & Winteler, A. (1992). Interest as a predictor of academicachievement: a meta-analysis of research. In K. A. Renninger, S. Hidi, & A. Krapp(Eds.), The role of interest in learning and development. Hillsdale, NJ: LawrenceErlbaum Associates.
Seidel, T., Prenzel, M., Duit, R., & Lehrke, M. (Eds.). (2003). Technischer Bericht zurVideostudie “Lehr-Lern-Prozesse im Physikunterricht”. Kiel: IPN.
Shek, D. T. L., & Ma, C. M. S. (2011). Longitudinal data analysis using linear mixedmodels in SPSS: concepts, procedures and illustrations. The Scientific WorldJournal, 11, 42e76. http://dx.doi.org/10.1100/tsw.2011.2.
K. Kiemer et al. / Learning and Instruction 35 (2015) 94e103 103
Sherin, M. G., & Han, S. Y. (2004). Teacher learning in the context of a video club.Teaching and Teacher Education, 20, 163e183. http://dx.doi.org/10.1016/j.tate.2003.08.001.
Sierens, E., Vansteenkiste, M., Goossens, L., Soenens, B., & Dochy, F. (2009). Thesynergistic relationship of perceived autonomy support and structure in theprediction of self-regulated learning. British Journal of Educational Psychology,79(1), 57e68. http://dx.doi.org/10.1348/000709908X304398.
Tripp, T., & Rich, P. J. (2012). The influence of video analysis on the process ofteacher change. Teaching and Teacher Education, 28(5), 728e739. http://dx.doi.org/10.1016/j.tate.2012.01.011.
Turner, J. C., Warzon, K. B., & Christensen, A. (2011). Motivating mathematicslearning: changes in teachers' practices and beliefs during a nine-monthcollaboration. American Educational Research Journal, 48(3), 718e762. http://dx.doi.org/10.3102/0002831210385103.
van Veen, K., Zwart, R., & Meirink, J. (2012). What makes teacher professional devel-opment effective? A literature review. In M. Kooy, & K. van Veen (Eds.), Teacherlearning that matters: International perspectives (pp. 3e21). New York: Routledge.
Walshaw, M., & Anthony, G. (2008). The teacher's role in classroom discourse: areview of recent research into mathematics classrooms. Review of EducationalResearch, 78(3), 516e551. http://dx.doi.org/10.3102/0034654308320292.
Webb, N. M. (2009). The teacher's role in promoting collaborative dialogue in theclassroom. British Journal of Educational Psychology, 79(1), 1e28. http://dx.doi.org/10.1348/000709908X380772.
Wells, G., & Arauz, R. M. (2006). Dialogue in the classroom. The Journal of theLearning Sciences, 15(3), 379e428. http://dx.doi.org/10.1207/s15327809jls1503_3.
Wilson, S. M. (2013). Professional development for science teachers. Science, 340,310e313. http://dx.doi.org/10.1126/science.1230725.
Recommended
